As a result, the proposed method, which is robust to predicted quality and variation of responses, yields better results than traditional desirability function method.

• 因此，用该方法进行优化能得到比传统满意度函数法较好的结果，且对模型的预测能力和响应的波动都具有稳健性。